Olivia:
“Bias” is a systematic deviation from a true, underlying population parameter. As such, the world cannot be biased, with respect to its policy preferences (nor can it have biased perceptions of these preferences), by definition. One could say that the mean preferred policies are overly liberal/redistributional, against a (say) neo-classical standard—but that is the point: that optimality, in this cause, would have an objective standard from which we might measure deviations.
By the same token, one cannot say that the media (or the academy) are biased, without some objective standard of “neutral”. It turns out that this is a devilishly difficult thing to do. It is not enough to say that because the members of the academy and the media are mostly Democratic (as has been shown by a number of studies), they will favor Democrats in their political coverage or classrooms. Simply “counting-up” coverage will also not get it done IN THE ABSENCE OF AN OBJECTIVE STANDARD OF WHAT THAT COVERAGE “SHOULD” BE. This is a missing link of some significance.
Tim Grosclose is likely among the most creative in attempting solve this riddle. A recent paper* is particularly clever: he uses content-analysis to connect members of the media to Congressional lib/con scaling measures—each of these measures and methods being more mature and (presumably) reliable than, say, self-identification. This paper finds great support for a liberal bias in the media—which should suit this audience great (here is the long-awaited “proof” of liberal bias, rigorously derived!).
A demand for precision is not an academic splitting of hairs. We should expect precision in elucidating our beliefs and in the empirical tests thereof; and we should demand no less from the various opinion-makers to whom we give audience. This is absolutely foundational to positivist (e.g. Lakatosian) research programs. Indeed, the “fuzzy” logic advocated by the post-modernist and ultra-liberal (and constructivist) camps, and so often derided here, is fundamentally inimical to such rigor. Unfortunately, imprecise statements, such as those by Olivia risk treading the same path—from the opposite direction.
*Tim Grosclose and Jeffery Milyo. “A Measure of Media Bias.” The Quarterly Journal of Economics. November, 2005: pp. 1191-1237.
